Lay health coaches provide support and accountability between health care visits to promote behavior change. There are 2 types of lay coaches, Peers (2 individuals initiating behavior change coach one another) and Mentors (individuals who previously and successfully modified their behavior coach patients just starting to change their behavior). Lay coaches have not been tested for obesity treatment. The only trials with objective outcomes are in the area of diabetes management, and results suggest that Peers may be more effective than Mentors. Similarly, findings from Social Comparison Theory show that lateral comparisons with similar targets (e.g., Peers) may promote greater motivation and health behavior change than upward comparisons with more successful targets (e.g., Mentors), as comparisons with successful targets can induce frustration and hinder performance. Although a natural fit, Social Comparison Theory has not been used to understand the effects of Peer and Mentor Health Coaches on health outcomes. Moreover, the2types of lay coaches have not been directly compared in any area. We conducted a 6-month pilot study testing Peer and Mentor Health Coaches for weight loss. Participants received a group behavioral weight loss program (BWL) and were randomized to a Peer or Mentor Health Coach. BWL treatment involved a reduced contact schedule; sessions were weekly for 6-weeks, biweekly for 6-weeks, and monthly thereafter. On weeks that group did not meet, participants emailed their week's weight, calorie, and activity information to their coach and received feedback. Results were impressive. Compliance with the coaching protocol was excellent. The Peer condition yielded superior weight loss outcomes compared to the Mentor condition (-8.7kg v. -5.3kg). Moreover, Peer participants engaged in lateral comparisons with their coach whereas Mentor participants engaged in upward comparisons, and lateral comparisons were associated with greater weight loss outcomes. These preliminary findings provide strong evidence that 1) lay health coaches are feasible for obesity treatment, 2) Peer Health Coaches may yield particularly impressive weight losses, and 3) social comparison processes may play an important role in Peer and Mentor Health Coach Treatment outcomes. We propose to test lay health coaches for obesity treatment. Participants will be randomized to BWL+Peer, BWL+Mentor, or BWL alone. BWL will involve a reduced contact schedule; sessions will be weekly for 6-weeks, biweekly for 6-weeks, and monthly for the remaining 9-months. During weeks that group does not meet, Peer and Mentor participants will email their progress to their coach and receive feedback. We hypothesize that BWL+Peer Coach will yield superior weight losses at 12-months compared to BWL+Mentor and BWL alone and that Peer and Mentor arms will differ on social comparison direction. If these hypotheses are supported, we will explore whether social comparison direction mediates 12-month weight loss outcomes between the Peer and Mentor study arms. This project tests a novel, theoretically-grounded, disseminable approach to obesity treatment. If successful, this approach could substantially inform current obesity treatment efforts.